This multicenter study, involving 16 Brazilian hospitals, investigated predictors of VTE in hospitalized patients with COVID-19 through logistic regression analysis and ML approaches. Among the variables identified by the multivariate logistic regression analysis as predictive of VTE, six were confirmed by ML analysis, which include D-dimer levels, axillary temperature, neutrophil count, C-reactive protein and lactate levels and SF ratio. D-dimer was one of the main predictors identified in both methods. The incidence rate of VTE was 6·7%, confirming the increased thrombotic risk in COVID-19 patients. Mortality, need for mechanical ventilation and renal replacement therapy were higher in patients who developed VTE in comparison with the patients who did not, highlighting the severity of this complication in the prognosis of COVID-19.
Our study is one of the largest individual studies on risk factors of VTE in COVID-19 patients. It confirms and extends the findings of a recent meta-analysis of 38 studies, which is the largest meta-analysis on the topic to date. This meta-analysis included 7,847 patients and showed that, among other factors, D-dimer, CRP and neutrophil count are independent predictors of VTE in COVID-19 patients. However, the results were limited due to high heterogeneity between individual studies. Overall, the number of studies assessing each clinical characteristic and laboratory tests was small, as well as the sample sizes, which may lead to the lack of power to detect associations we observed in the present analysis.(21)
D-dimer is an important marker of COVID-19 coagulopathy(6) and is often increased in patients with COVID-19, regardless of the presence of thrombosis. A previous study in 1,240 patients showed that the mean D-dimer in undiagnosed patients with pulmonary embolism was 1,371 µg/L.(19) Although D-dimer is also elevated in other conditions, such as sepsis, recent surgery and malignancy,(34) some studies have shown that D-dimer is an important predictor of VTE in COVID-19 patients.(9, 21, 25) A recent meta-analysis suggested that the traditional D-dimer cut-off value (<500 µg/L) used to exclude VTE in the general population seems applicable also to patients with COVID-19.(8) However, as a VTE risk predictor, there are still uncertainties about which levels would, in fact, predict a VTE. Additionally, the interpretation of D-dimer results is challenging due to the great diversity of methods, cutoff values, measurement units and whether presented as D-dimer units (DDU) or fibrinogen equivalent units (FEU), which are approximately twice those of DDU. A literature review has shown that the majority of studies that assessed D-dimer in COVID-19 patients did not make these points about the test clear, impairing the interpretability of the results.(35) Therefore, the analysis of D-dimer in relative values, compared to the reference value, seems to be more proper. A North-American study showed a significant increase in the strength of association the higher the levels of D-dimer,(13) while a Chinese study indicated that the most significant association with VTE occurred when D-dimer increments ≥ 1·5 fold.(36) In our study, the association of D-dimer with VTE was significant when D-dimer was four or more times above the reference value. A recent North American retrospective cohort observed the same cut-off for VTE or mortality.(25) These data suggest that this cut-off value may be a predictor of VTE in hospitalized COVID-19 patients. However, the previous study used maximum D-dimer levels during hospitalization, which may have been assessed at the moment of the suspicion of the diagnosis. As the main gap is to identify risk factors to help clinical decision, using D-dimer at hospital presentation seems more appropriate and more suitable for clinical application.
Our study showed other independent risk factors as predictors of VTE in COVID-19 patients, which were not previously identified in other studies,(21) such as recent surgery, being an ex-smoker, axillary temperature and lactate levels. Although some authors have questioned the role of traditional risk factors of venous thromboembolic disease as predictors of VTE among COVID-19 patients,(19, 37) our study reassures recent surgery and obesity as independent predictors.
Surgery has been consistently recognized as a major transient risk factor for VTE, among the general population, and recent data has shown that risk may remain elevated for 7 to 12 weeks after surgery.(38) It was quite unexpected that such association was not observed among COVID-19 patients in previous studies. We hypothesized that this may be due to the lack of power or to lack of collection of information on recent surgery in the previous studies. While obesity is considered a minor risk factor for VTE among the general population,(39) in COVID-19 patients, it has been shown to be associated with severe disease and increased risk of mortality.(13, 14, 19, 36, 40) Besides the association of obesity with venous stasis and decreased mobility, patients with obesity also have coagulation abnormalities, which may lead to increased risk of thromboembolic disease.(41–46) Increased plasma levels of fibrinogen, plasminogen activator inhibitor-1, factors VII and VIII, von Willebrand factor, increased platelet activation and higher circulating procoagulant microparticles as well as endothelial dysfunction have been reported in obese patients.(41–46)
Smoking is another minor risk factor for VTE in the general population.(47) It has not been observed to be a predictor among patients with COVID-19 in the more recent individual studies.(21) In our study, previous smoking was an independent predictor of VTE, but current smoking was not. This may be due to underreporting of current smoking, as the rate was less than 4%.
Unlike previous reports, our study identified axillary temperature upon hospital admission as an independent predictor of VTE risk. One of the hypotheses for this may be the consequent contraction of volume secondary to insensitive losses, contributing to the venous stasis of the Virchow's triad.(7)
In our study, inflammatory markers such as C-reactive protein and neutrophil count were independently associated with the occurrence of VTE, in agreement with other reports.(19, 21) However, unlike other publications, we also found that lactate level was an independent predictor of VTE in COVID-19 patients. Lactate level is a marker of disease severity and corroborates previous evidence that indicates an increased thrombotic risk in patients hospitalized with severe infections, such as sepsis and septic shock.(48)
Hospitalization due to acute infections has shown to be a strong trigger for VTE, independent of immobilization. It is hypothesized that this may be a consequence of the inflammatory response, leading to local activation of tissue, factor-mediated coagulation and to local vasoconstriction.(49, 50) In hospitalised patients with COVID-19, the cytokine storm, excessive inflammation and the consequent endothelial injury, inflammatory endotheliitis, besides hypoxia and disseminated intravascular coagulation are believed to play a key role in this process.(5, 7, 51)
In our study, we found that atrial fibrillation and flutter, SF ratio (peripheral oxygen saturation over inspired oxygen fraction) and prophylactic use of anticoagulant were protective factors for VTE, in hospitalized COVID-19 patients. The highest levels of SF ratio (peripheral oxygen saturation over inspired oxygen fraction) likely reflects a diminished severity of the inflammatory response. In fact, SF ratio was an important predictor of mortality in the ABC2-SPH score, derived from this same cohort.(52) Although less used than the ratio of arterial oxygen partial pressure over inspired oxygen fraction (PaO2/FiO2), SF ratio has been validated as a surrogate for the PaO2/FiO2 ratio, to assess the severity of hypoxemia, in patients with acute respiratory distress syndrome.(53) As it does not require vascular puncture, the SF ratio is less invasive, less painful and simpler than the PaO2/FiO2 ratio.
The findings of this study confirm those of a previous study which showed that pre-existing cardiovascular diseases are not associated with a higher PE risk, in COVID-19 patients.(19) However, in our study, the presence of atrial fibrillation or flutter was shown to be a protective factor of VTE. This is likely to be a proxy of anticoagulant use, since more than 90% of these patients in our study were using anticoagulants prior to admission and the vast majority of these patients had oral medication changed to therapeutic heparin during hospitalization. Despite this possible reduction in the rate of VTE with full anticoagulation, it does not mean that full-dose anticoagulation should be routinely administered to patients with COVID-19. A recent meta-analysis showed that the indiscriminate use of a full dose of anticoagulant significantly increased the incidence of bleeding and mortality.(54) On the other hand, a recently published randomized multiplatform trial indicated a potential benefit of routine therapeutic anticoagulation for patients hospitalized for non-critical COVID-19, in relation to days free of cardiovascular or respiratory organ support.(55) Another recently published randomized study not included in the meta-analysis(54) showed that the empirical use of anticoagulant at a therapeutic dose reduced the occurrence of thromboembolic events in patients hospitalized in a ward with D-dimer ≥ 4 times the reference value,(25) the same cut-off we observed as a predictor of VTE in the present study. More studies are still needed to better guide when and for whom to use the full dose of anticoagulant as a prophylactic strategy. However, our study corroborates the most recent evidence that a possible cut-off value of the D-dimer four times the upper limit of reference may be a guide for a more aggressive anticoagulation approach. As expected, in our study, the use of anticoagulants at a prophylactic dose reduced the risk of VTE in COVID-19 patients, corroborating data already available.(19)
In the present study, ML approaches detected other fourteen potential predictors of VTE in patients with COVID-19 in addition to the six variables identified by logistic regression analysis. One of the main advantages in traditional methods, such as regressions, lies in how simple they are and in how just analyzing the model (i.e. looking at the coefficients, for instance) can properly explain what was learned in the model.(32) Despite that, many of such techniques fall short in the sort of patterns they can learn, mostly remaining restricted to linear associations among variables, manually crafted non-linearities and other simpler variable associations. Additionally, LR's performance usually deteriorates in presence of collinearity, which may be especially problematic when the variables are not perfectly collinear and discarding some of them may result in useful information loss. Furthermore, missing values have to be replaced with some form of artificial values, which may also generate problems. Machine learning approaches have the ability of dealing with collinearity and redundancy, which may have occurred among some variables, as well as the ability to assess non-linear correlations.
Among the chief advantages of using ML models is their learning capacities, enabling them to capture much more complex patterns, sometimes even ascending into semantic and abstract levels. Albeit requiring substantially more data points in exchange. In the particular case of decision trees, random forests and gradient boosting machines, collinearity is not a problem, which means no potentially predictive information has to be discarded, and missing values do not require any form of filling.(56, 57) However, there is also an increased risk of identifying spurious (non-significant) associations, mainly due to issues of overfitting.(58)
In multivariate logistic regression analysis, we have not observed an association with some variables which were significant in the aforementioned meta-analysis,(21) including white blood cell count, alanine aminotransferase (ALT), lactate dehydrogenase (LDH), and prolonged prothrombin time, but these variables were observed as predictors in the ML model.
The rate of VTE in our study was at the lowest limit of that usually described in the literature.(9) Several factors may have influenced this finding. Some of the previous studies performed routine imaging exams or even excluded patients who had not performed imaging exams for VTE while asymptomatic for the disease, which may have overestimated the rate of thromboembolic events.(19) Additionally, the first cases of infection in Brazil only occurred at the end of February 2020, while the first wave of the disease only happened between April and May of that same year. At that time, the thrombogenic potential of the disease was already known and the routine use of thromboprophylaxis for patients hospitalized for COVID-19 was already widespread.(59) The rate of use of thromboprophylaxis, either at high or low dose, in our study was high, with more than 90% of the participants having used it. Therefore, we believe that this is the major reason for the lower incidence in our study. Furthermore, since the publication of the Recovery trial,(60) in June 2020, dexamethasone has been included in the treatment of patients with COVID-19, when they require using oxygen therapy or ventilatory support. It is possible that this has also influenced the decreased incidence of VTE, through reduced inflammation and, therefore, the thrombotic potential. There is already previous evidence suggesting that, in proinflammatory conditions, there is an increase in anticoagulant factors and a reduction in fibrinogen and procoagulant factors.(61) We also hypothesize there could be an underestimation of the occurrence of VTE due to limited access to objective tests, to avoid spreading out the disease. Nevertheless, even considering an incidence of 6·7%, it was higher than that described in other viral infections, supporting the thrombogenic potential of COVID-19.(62)
When compared to the group without VTE, the use of invasive mechanical ventilation, the need for renal replacement therapy and inhospital mortality were about twice as high in patients with VTE, reinforcing the prognostic importance of thrombotic events in patients with COVID-19.(13, 14, 19, 36, 40) As expected, the bleeding rate was higher in groups with VTE, due to the more frequent use of therapeutic doses of anticoagulants. However, most of these bleeding events were non-serious (Table 1). There was no difference in the severity of bleeding between the groups (Table 3).
This study has some limitations. First, the data collection was retrospective, which resulted in missing data on some laboratory tests. As a pragmatic study, they were requested at the discretion of the attending physician. In order to minimize this impact, we agreed, from the beginning of the study, that we would select, for data analysis, those centers with missing data accounting for less than 35% for D-dimer, as this is one of the potential VTE predictors at COVID-19 that most consistently appear in studies on the subject.(8, 9, 13, 36) Second, all variables analysed were collected upon hospital admission, as we would like to provide evidence to alert clinicians, so they could be able to identify, as soon as possible, patients at the highest risk of VTE, allowing for prompt diagnosis and treatment. Therefore, other relevant factors that could increase the risk of VTE, occurring during hospitalization, were not evaluated. Third, laboratory tests were not centralized. In particular, D-dimer was performed by using different methodologies, according to local hospitals. However, we evaluated D-dimer in relative values, stratified into five groups, in relation to the upper limit of normality of the reference values. Due to high variability in D-dimer assays among different institutions, we strongly believe that the way we analysed, increases the applicability of our findings.